Martin thanks for bringing up the topic. Some of the duplication was due to the refectoring effort that was done a while ago. If I am not mistaken the 0-10 client was using the URL proposed by the AMQP WG. That resulted in a different parser. Then it was again changed to the current URL format. (Arnaud correct me if I am wrong).
In general the refactoring resulted in code that is not used anymore. I agree that at some point it would be good to get rid of code thats not being used. There is plenty of methods and whole classes that are not used at all. Also I have seen duplicate methods that do more or less the same thing. As Martin pointed out it would be good if we could consolidate and eliminate duplicated as much as possible. Regards, Rajith. On Tue, Oct 7, 2008 at 11:38 AM, Martin Ritchie <[EMAIL PROTECTED]> wrote: > Hi, > Just looking at some of our latest commits reminded me of some of the > crazy things that happened to our code base pre-code review. Looking > at the duplication a bit just we have a number of duplicated > interfaces and multiple implementations that are used in a variety of > places > > BrokerDetails(one in qpid one in qpid.jms) interfaces and so two > implementations AMQBrokerDetails and BrokerDetailsImp? > > We also have THREE URL Parsers: > URLParser (client) > URLParser_0_10 (client) > QpidURL (common) and QpidURLImpl (common) > > Is there a reason for this duplication that I'm not aware of? > > Can we refactor these classes into single instances. I'm not even sure > we need to have interfaces here. > > I can just see this becoming a headache especially as we have a > BrokerDetails interface in the qpid.jms package that we have to > maintain from release to release. > > Regards > Martin > > -- > Martin Ritchie > -- Regards, Rajith Attapattu Red Hat http://rajith.2rlabs.com/
